Pearl Abyss acknowledges generative AI traces in Crimson Desert and promises to replace the content

Pearl Abyss acknowledged that the release version of Crimson Desert inadvertently included 2D materials showing signs of generative AI. After player complaints, the studio began an audit of files, added a warning on the game's Steam page, and promised to replace the disputed content in upcoming patches.

How the Problem Was Found

The controversy stemmed from player observations made immediately after the Crimson Desert release on March 20, 2026. Users noticed two-dimensional environmental elements — particularly paintings scattered throughout the game world — that displayed typical artifacts of generative models. This wasn't about disputes over overall style or narrative: players pointed to specific images that looked like quickly generated placeholders accidentally left in the final build of a major commercial game.

The situation was aggravated not by the mere presence of AI tools in early production stages, but by the fact that customers received no proper warning beforehand. For the audience, this appeared to be a hidden substitution of manual work in a product marketed as a major open-world fantasy action game. When such things are discovered after sales launch, the question quickly shifts from the quality of individual assets to trust: if the studio missed this in the release, it means internal quality control either failed or wasn't sufficiently transparent.

What the Studio Said

In an official statement, Pearl Abyss confirmed that player suspicions were correct. According to the company, the problematic materials were used in early development stages as supplementary references for finding the artistic direction, but were then accidentally included in the release version of the game. The studio separately acknowledged that this outcome did not meet its internal standards and took responsibility for it. This is an important moment: the developer didn't argue with the community over interpretations but publicly admitted the mistake and the need for corrections.

"This does not meet our internal standards."

At the same time, Pearl Abyss acknowledged a communication failure. The company stated that it should have more clearly informed the community about the use of AI tools, and shortly after, a warning about AI-generated content appeared on the Crimson Desert page on Steam. For the studio itself, this is not merely a question of storefront wording. In fact, it had to clarify where the boundary lies between temporary working materials for the team and what is acceptable to leave in a commercial release without separate disclosure to players.

What Will Change in the Game

Pearl Abyss is currently conducting a full audit of in-game files and preparing replacement content. The studio promised that AI-generated paintings will be updated in upcoming patches, and internal processes will be reviewed to ensure such materials don't reach the final version unnoticed. Essentially, this isn't just one cosmetic fix, but an examination of the entire asset production pipeline: from early artistic work to the final build that goes to PC, consoles, and other platforms.

• Full audit of in-game files
• Replacement of disputed images in upcoming patches
• Warning about AI-generated content on the Steam page
• Review of internal asset review and approval procedures

For Pearl Abyss, the situation is unpleasant also because it occurred immediately after a notable commercial launch. Crimson Desert was released on PC, Mac, PS5, Xbox Series X, and S, with first-day sales exceeding 2 million copies despite mixed Steam reviews at launch. When a game already has a large paying audience, even a localized dispute over a few assets becomes a reputational risk: players begin to look more carefully not only at patches and performance but also at how honestly the studio describes the content creation process.

What This Means

The Crimson Desert story shows that for players, the question no longer comes down to a simple "for" or "against" generative AI. The key criterion is transparency. If a studio uses such tools at the prototyping stage but then accidentally leaves their traces in the release without clear disclosure, this is perceived as a breach of trust. For the industry, the signal is simple: hiding such things becomes more expensive than establishing clear rules and quality control upfront.